Knowledge specification of an expert system

It is proposed that knowledge specifications be used as bases for developing and maintaining expert systems. It is suggested that through knowledge acquisition, a knowledge specification representing the kinds of knowledge and reasoning processes used to perform a task can be produced. A prototype can then be built to test and improve the knowledge specification. When a stable and satisfactory specification is obtained, a production system for end users, based on the specification rather than on the prototype, can be implemented. The knowledge specification guides system changes during maintenance. An experimental study to assess and improve this methodology is reported. Prototyping is discussed, an expert system knowledge specification is presented, and a methodology for creating a knowledge specification using conceptual structures is described. The methodology is compared with a currently popular methodology for expert system development. The proposal is primarily intended for medium- to large-scale expert systems, which may have several developers and whose users will not be developing the systems.<<ETX>>

[1]  Russ Abbott,et al.  Knowledge abstraction , 1987, CACM.

[2]  Donald A. Waterman,et al.  A Guide to Expert Systems , 1986 .

[3]  Victor R. Basili,et al.  Experimentation in software engineering , 1986, IEEE Transactions on Software Engineering.

[4]  Michael R. Genesereth,et al.  Logic programming , 1985, CACM.

[5]  Jesse F. Dillard,et al.  Expertise in assessing solvency problems , 1987 .

[6]  Imran A. Zualkernan,et al.  Specification for Expertise , 1987, Int. J. Man Mach. Stud..

[7]  Herbert A. Simon,et al.  The Sciences of the Artificial , 1970 .

[8]  Jon Doyle.  Expert Systems and the "Myth" of Symbolic Reasoning , 1985, IEEE Transactions on Software Engineering.

[9]  Dianne C. Berry,et al.  The problem of implicit knowledge , 1987 .

[10]  Paul E. Johnson,et al.  Contextual relevance in analogical reasoning: a model of legal argument , 1989 .

[11]  James R. Slagle,et al.  AGNESS: A Generalized Network-based Expert System Shell , 1986, AAAI.

[12]  James R. Slagle,et al.  A Method for Evaluating Candidate Expert System Applications , 1988, AI Mag..

[13]  Herbert A. Simon,et al.  Whether software engineering needs to be artificially intelligent , 1986, IEEE Transactions on Software Engineering.

[14]  David J. Mishelevich,et al.  Managing prototype knowledge/expert system projects , 1988, CACM.

[15]  Allen Newell,et al.  The Knowledge Level , 1989, Artif. Intell..

[16]  John F. Sowa,et al.  Conceptual Structures: Information Processing in Mind and Machine , 1983 .